Evolution of Social Behaviour Patterns in Primates and Man

Reprinted from PROCEEDINGS OF THE BRITISH ACADEMY 88 EVOLUTION OF SOCIAL BEHAVIOUR PATTERNS IN PRIMATES AND MAN W.G. Runciman, J. Maynard Smith, and R.I.M. Dunbar (editors) Oxford University Press Proceedings of the British Academy, 88, 119-143 Friendship and the Banker's Paradox: Other pathways to the Evolution of Adaptations for Altruism JOHN TOOBY & LEDA COSMIDES Center for Evolutionary Psychology, University of California, Santa Barbara, CA 93106, USA Keywords: reciprocity; altruism; co-operation; social exchange; reciprocal altruism; evolutionary psychology. Summary. The classical definition of altruism in evolutionary biology requires that an organism incur a fitness cost in the course of providing others with a fitness benefit. New insights are gained, however, by exploring the implications of an adaptationist version of the 'problem of altruism', as the existence of machinery designed to deliver benefits to others. Alternative pathways for the evolution of altruism are discussed, which avoid barriers thought to limit the emergence of reciprocation across species. We define the Banker's Paradox, and show how its solution can select for cognitive machinery designed to deliver benefits to others, even in the absence of traditional reciprocation. These models allow one to understand aspects of the design and social dynamics of human friendship that are otherwise mysterious. FROM A SELECTIONIST TO AN ADAPTATIONIST ANALYSIS OF ALTRUISM THEANALYSIS OF THE EVOLUTION OF ALTRUISM has been a central focus of modern evolutionary biology for almost four decades, ever since Williams, Hamilton, and Maynard Smith caused researchers to appreciate its significance (Williams & Williams 1957; Hamilton 1963, 1964; and Maynard 0 The British Academy 1996. 120 John Tooby & Leda Cosmides Smith 1964). The related concepts of conflict and co-operation have since developed into standard tools of evolutionary thought, and their use has transformed our understanding of everything from inter-organism interactions and kinship (Hamilton 1964) to inter-gene and within organism interactions and structures. For example, when applied to the genome these concepts lead straightforwardly to the derivation of the set of principles of intragenomic conflict that govern much about how genetic systems and intra-individual structures evolve (e.g., Cosmides & Tooby 1981). Indeed, pursuing the logic of conflict and co-operation has even led to a transformation in how biologists think of fitness itself-not just in the addition of kin effects to individual reproduction (Hamilton 1964), but also in the reconsideration of what entities it is proper to assign fitness to. It is clear now that sexually reproducing individuals cannot properly be assigned fitnesses, nor can they be correctly characterized as inclusive fitness maximizers, because the genome contains multiple sets of genes whose fitnesses cannot all be maximized by the same set of outcomes (Cosmides & Tooby 1981; Dawkins 1982; Haig 1993). For this reason, fitnesseb can only coherently be assigned to genes or sets of co-replicated genes rather than to individual organisms or groups. By this and other routes, the careful analysis of co-operation and conflict has led inexorably to the recognition that genic selection is the fundamental level driving the evolutionary process, with individual selection analyses as often inexact and frequently problematic oversimplifications. In this new world of biological analysis, folk concepts like 'self-interest' and 'individual' have no exact counterparts, and their uncritical use can lead away from the proper understanding of biological phenomena. There are two evolutionary pathways to altruism that have been proposed so far, kin selection, and reciprocal altruism. We think there are other pathways in addition to these two, and after revisiting the logic of reciprocal altruism we would like to explore several of them. Williams (1966) introduced the core of the reciprocal altruism argument, which was greatly expanded upon by Trivers (1971)' and fitted into the Prisoner's Dilemma formalism by Axelrod & Hamilton (1981; Boyd 1988). The argument is that altruistic acts can be favoured if they cause the target of the altruism to subsequently reciprocate the act. A population of reciprocating designs is stable against invasion by nonreciprocators if part of the design is the detection of nonreciprocation and the subsequent exclusion of nonreciprocators. This argument is, in fact, a transplantation into biology of the fundamental economic insight that self-interested agents can increase their own welfare through contingently benefiting others through acts of exchange, i.e., by exploiting the potential for realizing gains in trade, to use terminology from economics. The reciprocal altruism argument involves THE EVOLUTION OF ADAPTATIONS FOR ALTRUISM 121 the exploration of only one branch of the more inclusive set of logically possible exchange relationships-the branch in which there is a delay between the time at which the agent takes the altruistic action and her discovery of whether the act is contingently compensated. The natural category of exchange relationships and their timing and contingency is larger than this one line of analysis, and for this reason, we tend to term the more inclusive set of relationships social exchange. Classically, the analysis of the problem of altruism follows logically from its standard definition: An altruistic act is one that lowers the direct individual reproduction of the organism committing the act while simultaneously raising the direct individual reproduction of another organism (Williams & Williams 1957; Hamilton 1964; Maynard Smith 1964). Viewed in this way, an essential part of the definition of altruism is that the individual committing the altruistic act be incurring a diminution in its direct reproduction-that is, a cost. Altruism is not considered to have taken place unless such a cost is suffered, and the existence of this cost must be demonstrated before there is considered to be a phenomenon to be explained. With cost to direct fitness defining and limiting the class of instances of altruism, the explanatory task becomes one of finding a corresponding and greater consequent benefit to fitness, as when there is a sufficiently offsetting benefit to kin (Williams & Williams 1957; Hamilton 1963, 1964; Maynard Smith 1964). Although the definition of altruism is sometimes widened to include acts that are costly in terms of inclusive fitness, the definition remains cost-centered. As useful as this framework has been, we think that a modification in the classical definition of altruism may open the way to additional insights about biologically interesting social phenomena, particularly in humans. Before discussing this modification, however, it is necessary to review briefly the logic of adaptationism, because the two issues are tied together. To begin with, we think that some measure of confusion has been generated in evolutionary biology by failing to clearly distinguish the first level of evolutionary functional analysis, selectionist analysis, from the second level of functional analysis, adaptationist analysis (Williams 1966; Symons 1990, 1992; Thornhill 1991). The first is the widespread and often productive practice of analysing behaviour or morphology in terms of its current or even implicitly prospective fitness consequences. If used carefully, this can be a key heuristic tool, and its widespread adoption has contributed to the avalanche of functional insights achieved in the last forty years. However, just as individual selection analyses need to be reformulated into genic selection analyses to sidestep errors and accurately explain the full landscape of biological phenomena, so also selectionist models need to be reformulated into adaptationist analyses to capture more precisely the John Tooby & Leda Cosmides relationship between selection and phenotypic design (Tooby & Cosmides 1990a, 1992). Within an adaptationist framework, an organism can be described as a self-reproducing machine. The presence in these organic machines of organization that causes reproduction inevitably brings into existence natural selection, a system of negative and positive feedback, that decreases the frequency of inheritable features that impede or preclude their own reproduction, and that increases the frequency of features that promote their own reproduction (directly, or in other organisms). Over the long run, down chains of descent, this feedback cycle pushes a species' design stepwise 'uphill' towards arrangements of elements that are increasingly improbably well-organized to cause their own reproduction into subsequent generations, within the envelope of ancestral conditions the species evolved in. Because the reproductive fates of the inherited traits that coexist in the same organism are to some significant extent linked together, traits will be selected to enhance each other's functionality (with some important exceptions, see Cosmides & Tooby 198 1; Tooby & Cosmides 1990b for the relevant genetic analysis and qualifications). Consequently, accumulat- ing design features will often tend to sequentially fit themselves together into increasingly functionally elaborated machines for trait propagation, composed of constituent mechanisms-adaptations-that solve problems that are either necessary for trait reproduction or increase its likelihood within environments sufficiently similar to ancestral conditions (Dawkins 1986; Symons 1992; Thornhill

Load more